Human Rationality and Hursthouse`s Four Ends Introduction In the

Human Rationality and Hursthouse’s Four Ends
Introduction
In the last section of On Virtue Ethics, Rosalind Hursthouse addresses the question of
objectivity in virtue ethics. She is particularly interested in answering the following form of the
question: “Can we hope to achieve a justified conviction that certain views about which character
traits are the virtues (and which not) are objectively correct?” (Hursthouse 164). She argues that
we can. Her justification depends upon an analogy between animal goodness and human
goodness. She argues that animal goodness and human goodness share a common structure, and
that views concerning which character traits are virtues can be justified on the basis of this
structure (Hursthouse 224). She first develops the evaluation structure in relation to plants, and
modifies the structure as she “ascends the ladder of nature”, ultimately arguing that humans are
ethically good in respect of satisfying the same natural ends as the social animal (Hursthouse
223). The natural ends in question are survival, continuance of the species, enjoyment/freedom
from pain, and the well-functioning of the social group (Hursthouse 198). These ends, taken
together, constitute Hursthouse’s evaluation structure. The result is that if an animal, including
humans, satisfies the four ends in the characteristic way of the species, then the animal is to be
evaluated as good. Otherwise, the animal (including humans) is to be evaluated as bad.
In this paper, I assess whether Rosalind Hursthouse’s evaluation structure in On Virtue
Ethics is fit to reflect the nature of human and animal good. I argue that when Hursthouse’s
concept of “characteristic way” is identified with Michael Thompson’s concept of “natural
history”, an argument can be made that her evaluation structure is, in fact, warranted in context
of the non-rational social animals. But if the analogy is to carry over to the human ethical
context, however, then Hursthouse must show that “what [humans] rightly have reason to do” is
constituted by the successful pursuit of the four ends. I argue that in light of rationality’s ability
to “step-back” from its life-form, Hursthouse cannot justify this claim with the materials that she
provides.
The Two Senses of Naturalism
Hursthouse’s thesis that we can obtain a justified conviction concerning which character
traits are virtues is grounded in her particular brand of naturalism. For my purposes I will
distinguish between two types of naturalism in On Virtue Ethics. First, there is what I call
general neo-Aristotelian naturalism.
General neo-Aristotelian naturalism (GN) consists of two theses: (i) evaluations of
living things are species relative and (ii) living things are good in respect of living in the
characteristic way of the species.
Hursthouse shares GN with other neo-Aristotelians such as Philippa foot. Second, there is
Hursthouse’s particular brand of neo-Aristotelian naturalism. I call this Hursthouse’s neoAristotelian naturalism.
Hursthouse’s neo-Aristotelian naturalism (HN) is a particular way of developing GN;
it consists in the thesis that: satisfying the four ends in the characteristic way of the
species is the structure of good for both social animals and humans.
The four ends are (i) survival, (ii) continuance of the species, (iii) characteristic
enjoyment/freedom from pain, and (iv) good-functioning of the social group (Hursthouse 202).
The characteristic way is the way of the characteristically successful species member. 1 Taken
together, both non-rational social animals and humans are good in virtue of satisfying the four
ends in the characteristic way of the species. Hursthouse uses HN to justify her thesis that we can
obtain a justified conviction concerning which character traits are virtues. If HN fails, then so
does her ability to provide an objective criterion for distinguishing virtues from non-virtues. But
1
An organism can only be successful if it lives in the way that is appropriate to its species. For instance, a bear that
lives a successful life by the standard of survival and reproduction is nonetheless a defective bear if satisfies those
standards by eating trash instead of, let’s say, hunting.
before we evaluate whether HN can yield accurate evaluations of animal and human good, let us
clearer on what HN is committed to.
Hursthouse’s neo-Aristotelian Naturalism and the Agreement Condition
HN argues that human good and animal good share a common evaluation structure. Let’s
get clear on the animal context first. Hursthouse sums up the evaluation structure for non-rational
social animals (hereafter just animals) in the following passage:
So, summing up, a good social animal (of one of the more sophisticated species) is one
that is well fitted or endowed with respect to (i) its parts, (ii) its operations, (iii) its
actions, and (iv) its desires and emotions; whether it is thus well fitted or endowed is
determined by whether these four aspects well serve (1) its individual survival, (2) the
continuance of its species, (3) its characteristic freedom from pain and characteristic
enjoyment, and (4) the good functioning of its social group—in the ways characteristic of
the species. (Hursthouse 202)
In this passage Hursthouse is doing two things: 1) identifying the aspects that constitute an
animal’s goodness and 2) identifying the ends that determine whether a given aspect is good or
bad. Taken together, an animal is good in respect of being well-endowed with respect to its
aspects, and these aspects, in turn, are determined to be good or bad on the basis of whether they
well-serve the satisfaction of the four ends in the characteristic way of the species.
HN is a particular way of developing GN; what distinguishes HN from GN is that HN
develops an evaluation structure that makes important claims about the nature of the
characteristic way for animals. It makes two claims: 1) the four ends are the characteristic ends
for animals and 2) the aspects that are relevant to the satisfaction of characteristic ends (for
animals) are the animal’s parts, operations, actions, emotions, and desires. These claims
constitute an implicit conception of the characteristic way for animals, and it is this that entitles
the evaluation structure to make evaluations. Specifically, the implicit conception is as follows:
the characteristic way for animals is the satisfaction of the four ends via the specified aspects
(parts, operations, actions, emotions, and desires)--in the way that is characteristic of the species
(in question). Since the good just is the characteristic way, if Hursthouse’s evaluation does
indeed track the relevant features of the characteristic way, then Hursthouse’s evaluation
structure does indeed track animal good. It is also important to note that this implicit conception
of the characteristic way has the implication that a description of an animal can be given purely
in terms of its ends and aspects. I will not challenge this.
Now that we have a clear idea of what’s going on in HN, we can question whether the
two claims that constitute Hursthouse’s implicit conception of the characteristic way for animals
are correct. First, are the four ends really the characteristic ends for animals, and if so, are they
an exhaustive list? Second, are the aspects specified in the evaluation structure really the aspects
that bear on an animal’s satisfaction of the four ends, and if so, are they an exhaustive list? These
concerns point out two ways that Hursthouse’s evaluation structure can fail. Hursthouse’s
evaluation structure will fail if either the four ends are not really (or not the complete set of) the
characteristic ends for animals or if the specified aspects are not really (or not he complete set of)
the aspects relevant to the satisfaction characteristic end of the species. Thus, in order for
Hursthouse’s evaluation structure to be capable of reflecting animal good, it must be true that the
characteristic way of the animal just is the satisfaction of the four ends via the aspects (specified
in the structure)- in the characteristic way of the species.
We can now isolate a condition that must be satisfied if Hursthouse’s evaluation is to be
warranted in the animal case. I call an evaluation structure warranted if and only if it is capable
of reflecting the good in question. Thus the condition:
Agreement Condition (AC): The aspects that constitute an animal’s living in the
characteristic way are identical to the aspects that constitute an animal’s satisfying the
four ends in the characteristic way of the species.
If AC is satisfied, then the characteristic way just is the satisfaction of the four ends in the
characteristic way. AC is both necessary and sufficient for Hursthouse’s evaluation structure to
be warranted for animals. AC is a necessary condition because if it is violated, then the
characteristic way amounts to something other than satisfying the four ends in the characteristic
way. If there were such cases, then evaluation structure would be unable to track the good in
question. AC is a sufficient condition because the good in question just is the characteristic way.
Since the characteristic way, according to Hursthouse’s evaluation structure, just is satisfying the
four ends in the characteristic way, and the characteristic way is the good, an animal’s satisfying
the four ends in the characteristic way is sufficient for the good. In other words, if AC is true,
then it does not miss anything when it comes to animal good. Thus, AC is both necessary and
sufficient for a warranted application of Hursthouse’s evaluation structure.
Whether or not AC is satisfied is an empirical question. The empirical question can be
put in terms of the questions we asked above: Are the four ends really the characteristic ends of
the social animal? Are the specified aspects really the relevant aspects for satisfying the
characteristic ends? If the four ends are the characteristic ends of animals, and the aspects listed
in the evaluation structure are the relevant aspects, then AC will be satisfied. In such a case the
characteristic way just would be the satisfaction of the four ends in the characteristic way.
Hursthouse’s thesis that the four ends are objective criterion that can be used to determine
which human character traits are virtues is based on the analogy from animals where the
structure is first developed. If the evaluation structure is not warranted in the animal case, then it
will certainly not be warranted in the human case. So, in the following section I evaluate whether
there is reason to think that AC is met in the animal context.
Natural History, Characteristic Way, and the Ends Condition
Ultimately what is needed for a warranted application of Hursthouse’s evaluation
structure is an agreement between what the characteristic way is thought to be and what the
characteristic way actually is. Hursthouse needs to justify the claim that the satisfaction of the
four ends in the characteristic way is the characteristic way. This is certainly an open question,
for it is both logically and biologically feasible that satisfying the four ends (in the characteristic
way) is not the characteristic way for animals. Furthermore, now that modern science has
scrapped the Aristotelian view of life in favor of a Darwinian view of the life, many may even
deny that animals have a characteristic way at all. In this section I will do four things: 1) argue
there is good reason to think there is a characteristic way 2) elucidate how the concept of the
characteristic way is formed 3) set the conditions for the agreement between what the
characteristic way for animals is thought to be and what the characteristic way actually is and 4)
argue that Hursthouse satisfies this requirement in the context of animals.
Beyond what is implicit in her structure, Hursthouse does not offer an explicit account of
what the characteristic way is. In passing she says, “A species ‘characteristic way of operating,
etc’ is identified by how it characteristically operates in its natural environment, natural habitat”
[...]” (Hursthouse 203). So, for Hursthouse, the characteristic way is the way of going on of the
characteristically successful species member, and this is determined by looking at how members
of the species operate in their natural environment. This doesn’t give us much traction for
understanding where she got her evaluation structure, where animal ends figure in, and how a
concept of the characteristic way for a species is acquired. Even if we observe a certain species
and see that they statistically behave in a certain way, it may not follow that what is true of the
majority is also true the characteristic way. For instance, it might be that the majority is defective
(not successful), and therefore cannot inform our concept of what is characteristic of the species.
Thus, mere empirical observation doesn’t suffice to supply a true understanding of the
characteristic way. To answer the question of whether what we think of the characteristic way
(for a species) corresponds to the actual characteristic way (for a species) we must get clear on
what the characteristic way is. For this, I will turn to Michael Thompson’s work on life.
The crucial atom in Thompson’s thought on life is the “natural-historical judgment”.
Natural-historical judgments are expressions of the form “The S is (or has, or does) F”
(Thompson 64). S refers to the species in question and F refers to an aspect (part, operation, or
action) of the species in question. For instance, “The wolf has four legs” and “The wolf hunts as
part of the pack” are both canonical instances of natural-historical judgments. The complete set
of true natural-historical judgments for a species S is called the natural history of S (Thompson
72). That is, the complete set of true natural-historical judgments constitutes the complete
description of the species life-form, and this description is the natural history. 2 Thompson argues
that the natural history (of a species) can be “viewed as indirectly articulating the ideal, standard
or perfect operation of a bearer of this kind of life” (Thompson 55, AHF). 3 One might be
skeptical of talking about an ideal life-form for a species, but Thompson’s argument, as we will
see, is well founded.
Consider the following natural-historical judgment: The wolf hunts as part of the pack.
What does this sentence express? Before proposing his own account, Thompson considers four
2
3
Footnote on life-form.
The role of “indirectly articulating” is very important here. The natural history of a species does not express how
an animal should be set-up. That is, the natural history cannot be understood in terms of a normative analysis.
Rather, the natural history just is how a species’ life form is set-up, and thereby introduces a standard against which
the “goodness” of particular members of the species can be gauged. Put differently, the natural-history should not be
understood as expressing a normative claim, rather normative claims (about living things) should be understood as
implicitly employing natural-historical judgments.
ways of interpreting such a natural-historical judgment: 1) as a universal claim 2) as a statistical
claim 3) as a ceteris paribus claim and 4) as a normative claim. For my purposes, it is only
important for me to address the first two interpretations. 4
Interpreting our above natural-historical judgment as universal statement would amount
to the following: For every wolf x, if x is a wolf, then x hunts as part of the pack. 5 If this is what
natural-historical judgments express, then they are just simply false. For if they are just hidden
universal statements, then our natural-historical judgment could be defeated by one instance of a
wolf that does not hunt. We know there are such free-riding wolves. Clearly then, if we assent to
the truth of ‘the wolf hunts as part of the pack’, then it cannot be reduced to a simple universal
claim about wolves.
Interpreting our natural-historical judgment as a statistical statement would amount to the
following: The majority (90 percent, let’s say) of wolves hunt as part of the pack. This
interpretation does have prima facie appeal, but Thompson rejects this construal.
But, again obviously, although ‘the mayfly’ breeds shortly before dying, most mayflies
die long before breeding. And if the description of the ‘life-cycle’ of the monarch
butterfly told us ‘what mostly happens’, then it would soon be unnecessary to visit that
strange Mexican valley in order to wade knee-deep among them. (Thompson 68)
The important point made by this passage is that something can be statistically false of a species,
but nonetheless true of its natural history. This is certainly true of the mayfly and the monarch.
Thus, when giving a description of the life-cycle or rather (constructing a natural-history) for an
species, it is not merely statistical data that is taken into account; one must look at all of the
phases of the species’ life-cycle and the role that each aspect of the organism plays in the living
4
Explain.
Thompson regards this as a “plausible theory”, but says that “My purposes do not, I think, require that I refute it”.
(Thompson 67) I must admit that I do not quite understand why he would say this. Given my argument above such
an interpretation of natural-historical judgments would make all natural-historical judgments come out false.
5
of that life-cycle. This eliminates the possibility of construing natural-historical judgments as
merely statistical statements.
Although I will not discuss them in detail here, Thompson goes on to eliminate the other
possible reductions of the natural-historical judgment. If Thompson is right, then naturalhistorical judgments cannot be reduced to universal statements, statistical statements, ceteris
paribus statements, or reduced via normative analysis. Thompson ultimately concludes that they
represent a form of thought that Thompson calls “non-Fregean generalities”. In his own words:
“The dispiriting suggestion will be that the intended natural-historical judgments form a subclass marked off from the others by content and not by form” (Thompson 77). To understand this,
supply the following content to the form of the natural-historical judgment: “The car is red”.
Clearly this statement is logically reducible to the statement “For some x, x is a car and x is red”.
If Thompson is right, the when life enters in as the content to S is F, the statement becomes
irreducible to any other form of statement. This is what grounds Thompson’s conclusion that life
represents a logically distinct form of thought. Furthermore, if it is true that natural-historical
judgments cannot be logically reduced to any other form of thought, then perhaps there is
support for the claim that when we talk about life we are talking about something special;
namely, a natural history (or life-form).
But, if our attempt to articulate the natural history (the complete set of true naturalhistorical judgments) of a species is not just matter of what we statistically observe about a
species, then how is it decided what goes into the natural-history of a species? After all:
The attempt to produce a natural history, by contrast, expresses one’s interpretation or
understanding of the life-form shared by the members of that class […] My
understanding may of course be shallow or deep, extensive or narrow, mostly true or
largely mistaken. (Thompson 73, 2008)
This is precisely the problem that Hursthouse runs into: the problem of whether her concept of
the characteristic way for animals is the actual characteristic way for animals. 6 For although the
natural-history is the set of natural-historical judgments that are true of a species, it is a separate
question (our question) whether our interpretation of a species’ natural history is correct.
Thompson does not address this question, so I will now propose a possible way of answering. I
then argue that Hursthouse’s evaluation structure can satisfy AC in the animal case if she equates
her notion of ‘characteristic way’ with Thompson’s ‘natural history’.
Consider the mayfly again. Why is “The mayfly dies shortly after breeding” true of the
mayfly if it is not statistically true? That is, why do we attribute ‘dying shortly after breeding’ to
the natural history of the mayfly rather than ‘dying before breeding’? ‘The mayfly dies before
breeding’ is in no way a necessarily absurd natural-historical judgment. After all, there are some
animals that cannot reproduce (the liger) of which it is true that they die before breeding (if only
because they cannot breed at all). Perhaps the mayfly has an inner drive towards death, and it is
the defective mayflies the go on to breed. This construal of the facts would have the benefit of
aligning the natural-historical facts with the statistical facts, but we do not think it is true.
Biologically, it is highly implausible that an organism has no concern for its genes, so, what is it
that leads us to assign natural-historical judgments to a natural history if not statistical facts?
Natural-historical judgments are true of species because the referred to aspects play
important roles in the life cycle of the species. For instance, consider two natural-historical
judgments: ‘The wolf has a small freckle on its underbelly’ and ‘The wolf hunts as part of the
pack’. Even if the former judgment is statistically true of wolves, it is not likely to enter into the
natural history of the wolf. This is because having a freckle does not (let’s plausibly assume)
Note that Thompson does not classify beings according to their being non-rational social
animals, or animals, or so on.
6
play any role in the life-cycle of the wolf. On the other hand, we can imagine a number of
aspects that do play important roles in the life-cycle of the wolf, our latter natural historical
judgments is such a case. Thus, it appears that an aspect enters into the natural history of a
species only if the aspect plays an important role in the species’ life-cycle.
The aspects that enter in the natural history for a species is determined by assessing the
functional role that the given aspects play in the life cycle for the species in question. Now, what
is this functional role? The functional role is not relative to human interest; the natural history
(like the characterisic way) is a matter of what constitutes the good qua species in question.
Determining whether an aspect plays a functional role for the species qua species requires that
we determine the basic ends or aims of the species. Thus, what belongs to a natural history
comes down to whether the aspect in question serves the ends that matter to the species. For
instance, let us assign the end of “survival” to the mayfly. If we do this, then this rules out the
natural-historical judgment “The mayfly dies before breeding”, and supports the naturalhistorical judgment that we are likely to identify as true: ‘the mayfly dies after breeding’.
Recognizing the role that ends play in the construction of the characteristic way can help us
begin to see how aspects can enter into a natural history (for a species) while being statistically
false of the species as a whole. But it is important to note that the ends that are assigned (or
determined of) the species are just as much a matter of interpretation as whether it is true that a
certain aspect has consequences for that end.
Determining the ends of a species is the first crucial step in constructing the concept of
the natural history for that species. Without determining the ends of a species, the naturalhistorical judgments that enter into natural history will either be arbitrary or statistical. Neither
provides reason to think that the natural-historical judgment is true of the natural history for the
species in question. Thus, I argue (this is where I extend from Thompson) that the following
three-part necessary condition for a veridical construction of the characteristic way for the
species.
Ends Condition (EC):1) characteristic ends are assigned to the species in question. 2)
The characteristic ends, in part, determine what enters into the natural history of a
species. 3) The characteristic ends that are assigned to the species are the actual ends of
the species.
EC ensures that a natural history is constructed in light of ends of the species in question, rather
than by purely statistical observation (or otherwise). 1 of EC is necessary because without 1, 2
of EC cannot be satisfied. 2 of EC is necessary because true natural-historical judgments are
those that articulate the role that certain aspects play for the species’ life-cycle. The role is
fleshed out in terms of ends. Thus, a natural history that is not constructed with an eye to the
characteristic ends would fail where it is most crucial, recognizing the roles played by the aspects
of the species. The third condition is necessary because if the characteristic ends that one assigns
to a species are not the actual ends of the species, then the natural history would simply be false.
That is, if the characteristic ends that are assigned to the species are not the actual ends, then one
has simply mistaken what is important for the species.
As necessary conditions, EC helps us understand how the role that the ends of the species’
play a role in constructing the concept of a characteristic way. The natural history of a species is
the set of true natural-historical judgments. Our concept of the natural history is a set of naturalhistorical judgments. Whether our set of natural-historical judgments converges on the truth
depends upon whether we pick out the aspects of the species that play important roles in the lifecycle of the species. This requires picking out ends. What EC does is to ensure that the aspects
that enter into the natural-history indeed play important roles in the life-cycle of the species.
A natural history constructed in light of the ends condition is not, however, sufficient to
guarantee a complete match between a concept of the natural history of the species and the true
natural history of the species. This is because it does not defend against cases in which an animal
satisfies the four ends with an aspect that is uncharacteristic of the species. Consider a bear that
satisfies the end of survival solely by eating trash. It would obviously be incorrect to consider
“The bear eats trash” to be a true natural-historical judgment of the bear. EC, therefore, is not
sufficient to adjudicate between these cases. This does not, though, take away the value of EC as
a necessary condition though.
Now that we have established EC as a necessary condition for a veridical natural history
concept, I will argue that we can use it to justify Hursthouse’s evaluation structure in the animal
context. In order for Hursthouse’s evaluation structure to be warranted AC must be satisfied.
Correspondingly, the problem that we posed for Hursthouse is: why should we think that AC is
satisfied? That is, why should we think that the aspects that constitute an animal’s living in the
characteristic way are identical to the aspects that constitute an animal’s satisfying the four ends
in the characteristic way of the species? If we identify Hursthouse’s characteristic way with
Thompson’s natural history, then we can start to see why it might be likely. This entails
substituting Hurthouse’s concept of the characteristic way wherever Thompson’s natural history
occurs in EC. Before we proceed I should say a word about how Hursthouse’s work connects
with Thompson’s work. Hursthouse views herself as fleshing out a project of Philippa Foot’s in
Natural Goodness. In Natural Goodness, Foot borrowed extensively from Thompson’s “The
Representation of Life”. Thus, it is not implausible that Hursthouse would accept this
identification. Let us now see how EC might help her satisfy AC.
Replacing natural history with characteristic way only has consequences for 2 of EC. The
revised condition is: the characteristic ends, in part, determine what enters into the characteristic
way of a species. When we take 2 and 3 of EC together we can begin to see how the four ends
are related to the characteristic way. Namely, what enters into the characteristic way is in part
determined by the four ends, and the four ends are the characteristic ends of the animal. Thus, if
2 and 3 of EC are true, the reason why the four ends are able to evaluate animal goodness is
because the four ends, in part, determine the characteristic way. For argument’s sake, let us grant
the four ends are the characteristic ends of the non-rational social animal. After all, it appears to
be true that the activity of the social animals is steered in direction to survive, continue the
species, avoid pain/ characteristically enjoy, and contribute to the good functioning of the social
group.
How does this help AC? AC states that the following is necessary (and sufficient) for a
warranted application of Hursthouse’s evaluation structure: the aspects that constitute an
animal’s living in the characteristic way are identical to the aspects that constitute an animal’s
satisfying the four ends in the characteristic way of the species. If EC is true for Hursthouse, then
we can see why AC would be likely to be satisfied. According to 3 of EC, the four ends in part
determine which aspects (hunting as part of the pack or whatever) enter into the characteristic
way. Thus, the aspects that constitute an animal’s living in the characteristic way are,
themselves, determined by the aspects that well-serve the four ends. This shows a direct
relationship between the aspects that constitute an animal’s living in the characteristic way and
the aspects that constitutes an animal’s satisfying the four ends in the characteristic way. This
would increase the likelihood of AC being satisfied.
Now, it is important to ask what is missing. The four ends only in part determine the
characteristic way. The reason why was discussed with the bear example. It could be the case
that a certain aspect helps an animal satisfy the four ends even though that aspect is not part of
the characteristic way of the species, and thereby not good. EC cannot discriminate between
these cases. Well then, one might ask, “What is required to discriminate between these cases?
Doesn’t this just beg the question about what the nature of the characteristic way is”? This is a
hard question that is beyond the scope of this paper, but a short response is due. The way to
discriminate between cases of aspects that simply well-serve the four ends vs. aspects that wellserve the four ends in the characteristic way is to determine the unique way in which the species
solves problems concerning the four ends (their characteristic ends). This constraint cannot be
well-formulated as a condition without being somewhat question begging, but for our purposes,
all that matters is that EC eliminates the cases of aspects entering into the concept of the
characteristic way that do not actually play a role in the animal’s life-cycle (and thus are not
actually part of the characteristic way). Furthermore, we can see from EC that the four ends play
an active part in determining the characteristic way, which makes it likely that the four ends will
satisfy AC. In what follows I will present how Hursthouse completes the analogy and then see
whether EC can help out Hursthouse in the human case.
Making the Analogy to the Human Context
In the move from the animal context to the human context, rationality comes onto the
scene. Despite the introduction of this new aspect, Hursthouse argues that the structure of human
good retains the same structure as animal good; namely, satisfying the four ends in the
characteristic way. She briefly addresses the possibility of introducing a fifth end to
accommodate human rationality, but dismisses it, claiming that it is not clear what the fifth end
could amount to (Hursthouse 218). So instead of identifying another end for the human, she fits
rationality into the picture by assigning it the role of the characteristic way. The human
characteristic way, as Hursthouse defines it, is “a rational way”: “A ‘rational way’ is any way
that we can rightly see as good, as something we have reason to do” (Hursthouse 222).
According to this evaluation structure for human good, a good human is one that possesses
character traits that well serve the four ends in the human characteristic way, that is, in the way
that one sees he/she has reason to do.
Although there is a somewhat mysterious normative quality (the “rightly”) to the human
characteristic way, as far as the evaluation structure goes, Hursthouse has succeeded in making
the analogy from the animal context to the human context. The structure of animal good and
human good are identical; only the characteristic way has changed. She concludes that “human
beings are ethically good in so far as their ethically relevant aspects foster the four ends
appropriate to a social animal, in the way characteristic of the species” (Hursthouse 224). In the
same way that the inability to hunt makes a bad wolf, a human whose character traits cannot be
seen as fostering the four ends (as responsive to the demands of reason) makes a bad human
being. With the analogy made, Hursthouse is now in position to justify virtues from non-virtues
by appeal to an objective criterion (her evaluation structure). Specifically, the virtues are the
character traits that well serve the four ends in the human characteristic way.
Ends Condition in Humanity
Now we can ask whether Hursthouse’s evaluation structure is warranted in the context of
human ethical evaluation. In order for Hursthouse’s structure to be warranted in the human case
AC must be satisfied. Furthermore, we now have EC on the table that sets a necessary condition
on whether AC will be satisfied. If EC is to be satisfied, then 1) humans must have the four ends
as characteristic ends. 2) The four ends, in part, determine what enters into the characteristic way
3) the four ends constitute the complete set of (actual) characteristic human ends. If all of these
conditions are satisfied, then we have reason to believe that AC will be satisfied in the human
case.
The first thing that might be noticed in determining whether EC (with the four ends
instantiated) is true of humans is that the very structure of the human characteristic way is
drastically different than animal characteristic way. 7 If we understand Hursthouse’s characteristic
way in Thompsonian terms, the characteristic way for animal is the set of true natural-historical
judgments for the species. For Hursthouse, however, the human characteristic way is the
operation of one human aspect; namely rationality. Furthermore, while the animal characteristic
way is purely empirical, the human characteristic way (doing what we rightly see we have reason
to do) is thoroughly normative. These disanalogies make it difficult to see how EC might be
satisfied in the human case. This becomes particularly clear when we insert the substitute the
content of the rational way where human characterisic way occurs in EC:
1) Humans have the four ends as characteristic ends.
2) The four ends, in part, determine what humans rightly have reason to do.
3) The four ends constitute the complete set of (actual) human characteristic ends.
The problem most obviously comes with 2 of EC. How is it that the four ends in part determine
what happens at the level of reasons? In the animal case it was clear how the four ends
determined the characteristic way for an animal. We observe certain aspects displayed by the
members of the species, and the observed aspects became candidates for natural-historical
judgments only if the well-served the four ends. It seems true of animals that they are driven
towards the four ends. But in the human case, we might ask: why should we see how natures as
7
Hursthouse acknowledges this (Hursthouse 202).
reason giving? In the following section I use an argument from John McDowell to pose problems
for thinking that EC can be satisfied in the human case.
Rationality, Freedom and Second Nature
The following argument from McDowell seriously threatens the idea that human
rationality has particular defining ends. 8 McDowell begins the argument by imagining a wolf
that has somehow acquired rationality:
A rational wolf would be able to let his mind roam over possibilities of behavior other
than what comes naturally to wolves. Aside from the fact that it comes within the scope
of our pretense, that may seem obvious, and indeed it is. Even so, it reflects a deep
connection between reason and freedom; we cannot make sense of a creature’s acquiring
reason unless it has genuinely alternative possibilities of action, over which its thought
can play. (McDowell 170)
The rational wolf shows that there is a relationship between rationality and freedom; rationality
entails freedom. That is, the capacity of considering and entertaining alternative conceptions of
ways of going on forces one into the role of having to make decisions in the world in accord with
whatever representations one is opened up to. This is the condition that the rational animal faces.
A rational animal cannot excuse itself as “simply going along with what is going to happen
anyway” because the very condition that is acquired with reason is becoming unstuck from the
flow of events that previously just swept one along (McDowell 170). So, while it may be true
and normative of the non-rational wolf that “Wolves need to pool their energies, if their style of
hunting is to be effective”, the rational wolf can still question whether this fact really constitutes
8
“Two Sorts of Naturalism” (1995). John McDowell argues that there is a link between reason and freedom that
usurps the normativity of nature, and thereby poses an important problem for neo-Aristotelian ethical systems that
attempt to find a foundation in natural facts. Although Hursthouse does not purport herself to be providing a
foundation for ethics, the argument applies in the same way.
a genuine reason for himself to participate in the hunt (McDowell 171). “Why ought I to hunt
when I could get by perfectly well being a scavenger?” the rational wolf might ask.
McDowell’s thought experiment shows that the authority of the rational animal’s
characteristic way is not normative in the same way it is for the non-rational social animal. Let’s
apply this insight to our present problem. There is a direct connection between the four ends and
the characteristic way for non-rational social animals because the four ends play an active role in
constructing the characteristic way for non-rational social animals. In the rational animal case,
however, the direct connection is severed because the rational animal can step back from any
proposed standard of being. 9 Thus, we are left at a loss as to how 1 of EC can be satisfied. That
is, if rational beings can step back from their natures, and rationality itself does not entail a
commitment to any ends in particular, then what ground is there to suppose that the four ends are
the characteristic ends of the human? Furthermore, even ends are assigned to the human; it
would be false that what we have reason to do is determined by the four ends (violation of 2 of
EC). This is because rationality would first have to endorse them in order for them to be reasongiving. There is no sense in which the four ends acquire their normativity strictly from our being
social animals. It might be true that we are social animals, and that the four ends are the ends of
the social animal, but McDowell’s wolf shows that ends become normative for a rational animal
only if the ends are endorsed (or entailed) at a rational level. Of course, the ability to step-back
from any proposed standard does not show, in any form, that one’s form is no longer reason-
Hursthouse notes this disanalogy. “With the other animals, it is (almost) guaranteed that overall, there are good,
healthy members of the species, who, with a bit of luck, will be living well, for what counts as being good and living
well is determined by the standard that, so to speak, nature has laid down for them. But for us it is an open question
whether any human being is good, or living well, given what we could be, not something that has already been
determined by nature”(Hursthouse 228).
9
giving, but it does pose problems for the objectivity of what humans ought to find reason-giving.
And this is just the problem that Hursthouse is trying to answer.
Conclusion
If EC is going to help Hursthouse out in the human ethical context, then it must be the
case that the four ends constitute the standard of what humans rightly have reason to do. But as
we saw, if humans are like rational wolves, then there is a problem here. Humans can step back
from any standard. Unlike the animal case where there is a direct connection between the four
ends and the characteristic way, there is nothing inherently connecting “what we rightly see we
have reason to do” to the four ends. Thus, unless rationality somehow entails a commitment to
the four ends, there is no reason to suppose that Hursthouse will meet EC. That is, unless
rationality entails a commitment to the four ends, there is no reason to think that the satisfaction
of the four ends in a rational way just is the human characteristic way. Perhaps Kant can derive
substance from the form of reason, but Hursthouse does not provide the necessary materials.
McDowell’s rational wolf shows that despite our social animal being, our rational being can call
its normativity into question. So, although EC helps out Hursthouse’s evaluation structure in the
context of non-rational animals, it does not seem to help out Hursthouse in the human case.
Hursthouse’s claim that the virtues are the character traits that well-serve the four ends must be
justified in some other way.
Works Cited
Hursthouse, R. (1999). On Virtue Ethics. Oxford: Oxford University Press.
McDowell, J. (1998). Two Sorts of Naturalism. In Mind, Value and Reality
(pp. 167-197).Cambridge, Massachusetts: Harvard University Press.
Thompson, M. (2004). Apprehending Human Form. Royal Institute of
Philosophy Supplement, 54, pp 47-74 doi:10.1017/S1358246100008444.
Thompson, M. (2008). Life and Action: Elementary Structures of Practice and Practical
Thought. Cambridge, Massachusetts: Harvard University Press.
.